State Presence, Armed Actors, and Criminal Violence in Central America

ABSTRACT

This paper examines the relationship between the perceived presence of state forces and non-state actors and the levels of criminal violence affecting Central America. It contends that state presence in some communities does not necessarily lead to lower crime levels than communities where state institutions are absent. The data of this study come from three nationally representative surveys conducted in Guatemala, Honduras, and El Salvador. The findings suggest that state forces, such as the police and the military, have a limited role in reducing criminal violence rates. In contrast, non-state armed actors, such as vigilantes and community groups, may have a more significant role in reducing violence in some circumstances. We analyze these findings as evidence of the complex relationships between state actors, non-state actors, and criminal violence in developing countries.

KEYWORDS:

• Crime
• state forces
• non-state armed actors
• Central America

Acknowledgments

This paper is part of a research project on policing in Central America, funded by the Open Society Foundations (Grant Number OR2015-22547). The authors are grateful to David Holiday for the support to this research. They also thank Angélica Durán-Martínez, Frank Mora, Jonathan Rosen, María Micaela Sviatschi, and the anonymous reviewers for their helpful comments on this article. An earlier version was presented at the Conference “Criminal Governance in the Americas,” at the University of Chicago on October 26, 2018. The final responsibility for the content is, of course, solely of the authors.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1. Information regarding the Latin American Public Opinion Project (LAPOP) and the methodology of the AmericasBarometer can be found at http://sitemason.vanderbilt.edu/lapop/about_us.

2. We recoded this variable into two groups and conducted a multivariate logistic analysis instead of an ordered logistic regression with the original ordinal variable because the parallel assumption was violated. Besides, a logistic regression facilitates the interpretation of the results.

3. The data for homicide in El Salvador come from: http://fundemospaz.org.sv/informes/estadisticas-por-municipio/homicidios/ and Instituto de Medicina Legal (2019). In Guatemala, they come from the Dialogos, a think tank that runs a public observatory of security (see www.dialogos.org.gt). In Honduras, homicide data correspond come from the National Honduras University’s Instituto Universitario en Democracia, Paz y Seguridad (see http://app.iudpas.unah.edu.hn/participacionciudadana/Denuncias/mapa_oficial) and from Honduran Secretaria de Seguridad: https://www.sepol.hn/sepol-estadisticas-registro-fallecidos.php. We thank all institutions for providing access to their data.

4. See the Appendix for a visual representation of the distribution.

5. When we refer to street gangs in the analysis, we refer to the item’s results that asked directly about the presence of gangs, which is different from “other” security-related groups.

6. The socioeconomic stratum is a variable constructed based on the distribution of monthly family income in Guatemala and Honduras. In El Salvador, the survey did not include the question of family income but monthly family expenditures. Hence, the socioeconomic stratum was constructed using that item.

7. The HDI for each municipality in the Northern Triangle was obtained from the last UNDP’s report that estimated the index at the municipal level.

8. See note 3 above for the sources of these data.

9. We used Likelihood-ratio tests for over-dispersion to confirm that negative binomial regressions were the proper technique. They produced a significance level of less than 0.001, ratifying the approach’s adequateness instead of Poisson regression.

Additional information

Funding

This work was supported by the Open Society Foundations [OR2015-22547].

Notes on contributors

José Miguel Cruz

José Miguel Cruz is director of research at the Kimberly Green Center for Latin American and Caribbean Studies (LACC) at Florida International University (FIU) in Miami. Previously, he was director of the Institute for Public Opinion (IUDOP) at the Central American University (UCA) in San Salvador. His areas of research are criminal violence, gangs, police, democratization, and public opinion in Latin America.

Yulia Vorobyeva

Yulia Vorobyeva is a former Postdoctoral Research Fellow at Florida International University’s Kimberly Green Latin American and Caribbean Center, where she conducted research on gang violence and the rule of law in Central America. She teaches graduate courses on Latin American Studies at Florida International University.